Ice-core drilling at Site A, Greenland (70°45′N, 35° 57.5′W; 3145m a.s.l.) by the Polar Ice Coring Office recovered two cores of about 100 m each, which are now in storage awaiting further analysis. Field studies show that visible strata in the cores record both individual storms and individual years; a typical year comprises 15 storm deposits. Pore close-off occurs between 75 and 80 m depth. The grain-growth rate in isothermal firn and ice is about 7.8 × 10−3mm2a−1, the 10 m temperature is about –29.5°C, and the accumulation rate over the last 200 years is about 290 mm a−1 ice. Melt events occur about once every 40 years, and major events can be correlated between cores separated by 20 m.

Concentrations of lead (Pb) have been measured by the ultra-clean isotope dilution mass spectrometry technique in various sections of the Antarctic Dome C and Vostok deep ice cores, whose ages range from 3.85 to 155 ka B.P., in order to assess the natural, pre-human, sources of this toxic heavy metal in the global troposphere. Pb concentrations were very low, as low as about 0.3 pg Pb/g during the Holocene and probably during the last interglacial and part of the last ice age. On the other hand, they were quite high, up to about 40 pg Pb/g, during the Last Glacial Maximum and at the end of the penultimate ice age. Wind-blown dust from crustal rock and soil appears to be the main natural source of Pb in the global troposphere. Pb contribution from volcanoes is significant during periods of low Pb only. Contribution from the oceans is insignificant.

The results of the 1984-85 post-GISP campaigns in central Greenland are presented. Eight ice cores were obtained, some spanning up to 360 years. We present:

• I. Geographical positions and elevations at the drill sites,

• II. Density and temperature in the bore holes,

• III. Filtered δ18O profiles and accumulation-rate variations along the ice cores.

The δ18O and accumulation profiles, along with those from the 400 m Crête ice core obtained in 1974, are compared.

The accumulation-rate series over a large region have a very high coherence. This indicates that a single bore hole in this region would give a representative accumulation time series over a long period. Further, it appears that for this region there is a linear relation between the measured annual accumulation rate and the mean annual δ18O values. Thus the area between Crête and Summit, just west of the ice divide, seems to be favorable for a deep-drilling operation.

Major volcanic eruptions deposit large amounts of strong acids in polar ice. Two such volcanic eruptions are Laki, A.D. 1783, at high latitude (64 °N), and Tambora, A.D. 1815, close to the Equator (8°S). The acid ice layers from these eruptions are easily reached by shallow drilling, and the acidity of the ice cores obtained has been determined by a solid electrical conductivity method (ECM), and in some cases by liquid pH measurements. The strong acid is identified by chemical anion analysis. Sulfate is the dominant anion in both of these volcanic events.

Atmospheric thermonuclear-bomb tests ejected radioactive debris into the atmosphere. Two major groups of such tests are those carried out by the Americans in 1952-54 at low latitude (11°N) and by the Russians in 1961-62 at high latitude (75°N). Radioactive debris from these events was deposited in polar snow, and can be detected by specific total β activity measurements. The radioactive layers serve as dating horizons in the firn. The total β activities were measured at least 10 years after ejection, thus the measured activities were mainly due to 90Sr and 137Cs.

The amount of 90Sr and 137Cs ejected into the atmosphere is known. We assumed a similar global distribution pattern of bomb-produced total β activity and strong acids from violent volcanic activity, and were able to calculate that both major volcanic events produced some 300 million tons of sulphuric acid. This is in agreement with other estimates of the Tambora eruption, which are based on studies of ice cores from Antarctica.

In 1978 two ice cores were drilled to depths of 46 and 92 m respectively at Camp 3, at the west margin of the Greenland ice sheet. Both core drillings reached bedrock. In addition, surface samples were collected in the marginal area along an estimated flow line.

The δl8O profiles of the two ice cores and of the surface samples show similar features. All three δ180 records reveal the characteristic shift (of 5–6 per mil for the Pleistocene-Holocene transition 11 000 years ago) observed in Greenland deep ice cores from Camp Century and Dye 3.

The δ18O results, as well as the measured temperature profiles in the bore holes, are used to provide more insight into the rheology of the ice sheet. The analyses of marginal ice samples is an important supplement to deep ice-core analyses.

A technique for extracting and analysing large air samples from bubbles occluded in an Antarctic ice core is discussed. Core samples of up to 1400 g were milled to release approximately 120 cm3 of air, which was dried, collected in a cold finger and then analysed by gas chromatography. The concentrations of atmospheric carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O) over the past 450 years have thus been revealed. Measurements of a chlorofluorocarbon (CCl2F2) in the ice-core air were used to check core quality and the air-occlusion process.

The ice core, designated BHD, was thermally drilled from the summit of Law Dome, Antarctica, where the average accumulation rate is 0.65 m a-1 water equivalent and the annual average temperature is –22°C. Ice dating was achieved by counting annual cycles of oxygen-isotope ratio and d.c. conductivity, and air dating was deduced from the density profile.

The results show the pre-industrial concentrations of the gases to be 288 ± 5 ppm volume for CO2, 800 ± 50 ppb volume for CH4 and 285 ± 10 ppb volume for N20.

Wind plays an important role in determining accumulation and δ(18O) on some ice caps. Three surface-to-bed cores spaced about 1 km apart have been taken on a flow line of the Agassiz Ice Cap, Ellesmere Island. The A84 core comes from the top of a local dome. The A79 core is 1200 m down the flow line, but very close to the ridge through the local dome. The A77 core is 1100m from A79 and well away from the ridge. The ridge causes wind turbulence, which removes or scours the soft winter snow from the A84 and A79 sites. No snow is scoured from the A77 site. Because of scour the retained accumulation and average δ(l8O) are different. The accumulations are 17.5, 11.5, 9.7 cm/a (ice equivalent) at A77, A79 and A84 respectively and the corresponding surface δs are –30.40, -27.90 and –27.05‰. The core records were dated by annual layer thicknesses and by identification of electrical conductivity measurement (ECM) acid peaks. With the three cores accurately aligned we examine the (δA84-δA77) and (δA84-δA79) time series. Significant variations in these difference series are interpreted as being caused by changes in the seasonal δ amplitude, which is then explained by changes in sea-ice cover. A seasonal δ amplitude series independently obtained from the Devon Island ice cap δ noise record is consistent with that from the Agassiz Ice Cap sites.

The present paper gives the preliminary results of the analyses on microparticle concentration and electrical conductivity of a 700.56 m ice core from Mizuho Station, Antarctica. Concentration of microparticles coarser than 0.63 μm in diameter increases more than twofold at the 240-440 m depth interval compared with that below 440 m in depth. The higher particle concentration is well associated with higher electrical conductivity and lower δ18O. Periods of high particle concentration are estimated to be 3000-6000 years B.P. A visible volcanic dirt band was found at 500.7 m below the surface. This dirt band may be isochronous with the shallowest ash band of the Byrd Station core, found at 799 m depth. The present study indicates that large-scale environmental changes possibly occurred in the Southern Hemisphere in the middle of the Holocene.

A 100 m ice core from the Ronne Ice Shelf, drilled during the 1983-84 field season, was dated by isotopic stratigraphy, using the well-known seasonal variation in the 18O content in firn and ice; the layers at a depth of 89 m are probably 400 years old. Layer thicknesses deduced from the 18O profile indicate short-term variations of the snow-accumulation rate over the last 400 years. The area of deposition of the material recovered with the core is estimated by a two-dimensional flow model and by the 18O content of the core, which decreases from –27.5‰ in the upper part of the core to –32.0‰ at 89 m depth.

In order to investigate the deposition of radioactive fission nuclides in the high-altitude snow and firn of the Swiss Alps after the Chernobyl accident, core samples from nine mountain peaks were analyzed. The observed radiological signal is described, together with the characteristics of the snow in which it is embedded. Total activity is of the same order of magnitude (around 10 kBq/m2) as deposition rates measured at lower altitudes, and shows a comparable regional distribution. A clear altitudinal trend only exists with respect to snow conditions: above about 4000 m a.s.l., erosion, mixing and resedimentation of cold snow by wind could have occurred during or shortly after the deposition, whereas at lower altitudes formation and percolation of melt water had already started — probably selectively — to redistribute the nuclides within temperate firn layers.

Preliminary results of the analyses on 700 m ice cores retrieved from Mizuho Station, Antarctica, in 1983 and 1984 are presented. The majority of the physical properties, density, grain-size and shape, and total gas content, were measured at the drilling site. Fabrics, microparticle concentration, electrical conductivity, and stable-isotope concentration δ18O were measured in laboratories after the cores had been taken to Japan.

In spite of inaccuracy in measuring both density and total gas content in the ice, due to interlocking cracks in cores, several attempts were made to correct the data. The coincidence between the incremental peaks in the depth profile of the microparticle concentration, as well as in the electrical conductivity and the warm trend indicated by the δ18O profile is discussed. The shape of the δ18O profile is characterized by two inflection points and is compared with results obtained from the Byrd Station, Dome C and Vostok cores. From this comparison, it is tentatively concluded that the bottom of the Mizuho core may be an age of the order of 10 ka B.P.

An historical record of the deposition of common acids is contained in snow and ice cores taken from suitable sites in the accumulation zone of certain glaciers. Spatial and time-series data sets for trace-mineral acids have been obtained from snow-pit samples and ice cores from a number of mountain sites in Alberta, British Columbia, and the Northwest and Yukon Territories. In Alberta, it is possible to use temperate firn sites above 3460 m, although elution occurs during certain summers as indicated by isotopic and ionic data. This would also apply to sites of a similar latitude (52°±2°N) in British Columbia. In the Yukon Territory (≥60.5°N) reliable time series for the acid anions may be obtained from sites at altitudes above 3000 m. Elution provides a natural control for demonstrating that field sampling and subsequent analytical procedures do not introduce significant contamination. The Yukon data are compared with the net annual accumulation rate and with altitude. Recent data from the 5340 m Mt Logan site do not indicate any significant increase in natural background levels of snow acidity. Lightning, which is responsible for numerous forest fires in all provinces, is a possible natural source of nitric acid. Spring-summer peaks in nitrate concentration usually occur. In addition, forest-fire smoke may be a significant contributor to the mountain snow-pack chemistry in some years and must be considered when interpreting the Mt Logan core data. One Yukon profile seems to contain the signature from the 1986 Augustine volcanic eruption.

The salinity and isotope (18O, 3H) content of multi-year landfast sea-ice (MLSI) cores from northern Ellesmere Island, Canada, are examined. Salinity ranges from 0.01‰ to 4.54‰, and δ18O ranges from −23.8‰ to +0.7‰. Salinity and δ18O are linearly related, and tritium values generally exceed natural background levels. The results are evidence of ice growth associated with fresh-water / sea-water stratification below the ice. Salinity variations are cyclic and indicate a mean annual bottom accretion rate of 0.33–0.5 m a−1. Rather than signifying downward percolation of melt water from the surface, the ice δ values are a proxy measure of variations in salinity and 18O content of the water below the ice. Annual salinity layers are preserved in the absence of significant brine movement and ice deformation. The fast-ice environment appears to favour the maintenance of water stratification and growth of annual layers. It is suggested that ice growth in this environment is somewhat independent of thermodynamic sea-ice growth models; instead, ice growth by a double-diffusion process might account for the growth of MLSI beyond thicknesses normally encountered in undeformed multi-year pack-ice floes.

Ice-core drilling and ice-core analysis (electrical conductivity–salinity, 18O, 3H, density) reveal that the internal structure of the west Ward Hunt Ice Shelf contrasts sharply with that of the east ice shelf. The west ice shelf contains a great thickness (≥22 m) of sea ice (mean salinity, 2.22‰; mean δ18O, -0.8‰), whereas the east ice shelf is entirely of meteoric or fresh-water ice (mean salinity 0.01‰; mean δ18O, -29.7‰). High tritium activities are found only in ice from near the bottom of the east and west ice shelves. The contrasting ice-core data is considered to be a proxy record of variations in water circulation and bottom freezing beneath the ice shelf. The west shelf is underlain by sea water flowing into Disraeli Fiord. Sea ice accretes on to the bottom of the west ice shelf from the sea-water flowing into the fiord. Sea-water flowing out of the fiord is directed below the east ice shelf. However, the east ice shelf is not underlain directly by sea-water but by a layer of fresh water from the surface of Disraeli Fiord. In this region, ice growth resulting from the presence of this stable fresh-water layer has been accompanied by surface ablation over a period of perhaps the last 450 years. As a result, fresh-water ice has completely replaced any sea ice that originally grew in the region of the east ice shelf. Whereas the west and east shelves are underlain almost exclusively by sea-water and fresh water, ice in the south shelf is the result of freezing of fresh, brackish or sea water. This is attributed to mixing of the inflowing and outflowing waters.

We analyzed ice cores from both northern and southern polar regions to determine the concentrations of nitrous oxide in the pre-industrial and ancient atmospheres from about 150 years to 3000 years B.P. We found that the pre-industrial concentration of nitrous oxide remained constant over the period we studied and that the average atmospheric concentration was 285 ± 1 ppb volume (90% confidence limits), representing about 2100 Tg (2100 × 1012g) of N20 in the atmosphere, whereas the average concentration in 1984 was about 307 ppb volume or 2260 Tg. This is a change of 22 ppb volume (160 Tg), or about 8%, between pre-industrial and present times. Now the rate of change is between 0.7 and 0.9 ppb volume/year or 5 and 6.5 Tg/year, which is a slow increase of about 0.3% per year. The changes observed are probably caused by increasing use of fossil fuels, particularly coal and oil, and perhaps to a lesser extent by the use of nitrogen fertilizers in recent years. The atmospheric lifetime of N2O is probably between 100 and 150 years. The pre-industrial concentrations, present levels, and a lifetime of 100 years are consistent with natural sources, mostly soils and oceans, of about 22 Tg/year and the present anthropogenic sources of about 8.7 Tg/year. In the next 50 years we expect nitrous oxide levels to reach 360–390 ppb volume, or about 16–25% more than present.

Major soluble chemical impurities have been measured along a 130 m firn core from the Amundsen–Scott Station in order to assess Southern Hemisphere environmental variability over the last millennium. Particular attention is given to the possible impact of the Little Ice Age, a well-known climatic disturbance which occurred in the Northern Hemisphere between about A.D. 1500 and 1900.

Na+, K+, NH4+, Cl+, SO42− and NO3− concentrations were carefully determined in forty-two 40 cm firn sections. Stringent precautions were taken to ensure the analytical reliability of the data set obtained. The average concentrations are (in ng g−1): 11.0 ± 2.5, 0.7 ± 0.4, 0.5 ± 0.2, 31 ± 5.6, 58 ± 11.6 and 103 ± 11.6 respectively (the scatter represents the standard deviation).

No definite trend is detected which could be linked to the Little Ice Age disturbance.

Previous pollen analyses of ice cores from Devon and Ellesmere islands have contributed considerably to our knowledge of past climate in the Canadian High Arctic. In this case, in 1979, bulk (35–83 litres) water samples were melted down a hole 139 m deep, drilled to bedrock, 1.2 km from the top of the flow line in Agassiz Ice Cap in northern Ellesmere Island. Analysis of ten of these samples, plus some taken in very dirty ice from the melt tank during drilling 7 years ago, has yielded pollen concentrations that, together with the oygen-isotope (6) signatures, suggest the Agassiz Ice Cap began its growth during the last interglacial period. A discrepancy between melt-tank and bulk-sample pollen concentrations is believed to be due to a loss of pollen from the melt-tank samples during the drilling process.

An integrated system for ice-fabric analysis on a Rigsby stage is described. The system consists of a regular Rigsby stage fitted with two opto-electronic sensors for assessment of azimuth and the tilt angle of each individual grain. Signals from the sensors are transmitted to a computer terminal via an interface box, which facilitates transformation of Gray-coded data to ASCII data records. The terminal is hooked up to a main-frame computer (VAX 750), where the digitized angles of the c-axis orientations of individual thin sections are stored in separate data files. These files are compatible with other already existing files containing additional ice-core data and thus become part of an extensive data bank. Appropriate software has been developed to produce, among other things, plots of c-axis orientations in a Schmidt net.

Sea ice constitutes a major element in the atmospheric, oceanographic and biological regime of the polar regions. Assessment of its fundamental properties requires interdisciplinary investigations on local, regional and global scales. Sea-ice structure and textural parameters of individual ice cores play a key role in such investigations. A proper characterization of sea-ice micro-structure is essential for an adequate classification of ice cores, an understanding of the growth processes of the sampled floe, and the identification of possible relationships between ice texture, and the physical, chemical and biological properties of sea ice. Investigations on ice cores which were obtained during three recent Antarctic expeditions (1983–85) in coastal waters of the eastern and southern Weddell Sea are reported. The basis for a number of physical, chemical and biological investigations is an assessment of the textural characteristics of each sea-ice core. These are derived by inspection of continuous thick sections, supplemented by an analysis of selected vertical and horizontal thin sections. Major results of this study can be summarized as follows: (i) in addition to the common ice classes, another sea-ice type, platelet ice, is identified; it is apparently unique to the coastal waters of Antarctica, near the ice-shelf edge, and (ii) different physical, chemical and biological sea-ice properties vary systematically with and are probably related to / controlled by the ice texture of the cores.

A strong volcanic-acid signal is clearly registered, using an acidity-measuring technique, in the A.D. 1259 ice layer in four different Greenland ice cores (Camp Century, Milcent, Crête and Dye 3). This signal is similar in amplitude to the Laki (Iceland) A.D. 1783 volcanic event as recorded in the central and south Greenland ice cores. Measurement of ice layers from corresponding age levels in Antarctic ice cores (Byrd Station, South Pole and J–9 on the Ross Ice Shelf) provides similar strong acid signals. There is no historical record of a significant volcanic eruption for the period around A.D. 1260 in the Northern Hemisphere. Subsequent chemical analyses of all A.D. 1259 ice layers show similar compositions. We suggest that the A.D. 1259 signals registered in both Greenland and Antarctica were caused by the same volcanic disturbance and that its epicenter was located at the Earth’s equatorial zone, which enabled global distribution of the acid gases. These results indicate that inter-hemispheric dating of ice sheets is possible by the chemical identification of major eruptive volcanic events in the equatorial zone.